• State Resolved
  • Locked Locked
  • Replies 17 replies
  • Subscribers 64 subscribers
  • Views 667 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

UCC3818: high frequency oscillations at light loads

Gabriele Montecchi
Prodigy 100 points
Part Number: UCC3818

I used UCC3818 in a 1KW power supply converter from 230Vac, no full range. At no load, or very small load, the VEAOUT signal is very low, about 0.3V and some pulses are missed, but is normal. The issue is that each pulse oscillates at a high frequency of some MHz, that I found at pin 16 output. I suppose the zero power comparator as a role in this behaviour because there is no hysteresis. Do you agreed and have some ideas to overcome this isssue?

Thranks

Best Regards

GMontecchi

  • 0
    TI__Mastermind 34465 points
    Hello Gabriele,
    What happens is that when the zero power comparator goes active it bypasses the PWM section and turns off the gate drive.
    This causes the current amplifier to saturate and when the zero power amplifier goes inactive again the gate drive comes on at full duty cycle.
    This may cause loop response issues, OVP or audio noise as the loop goes from max duty cycle to off . You should see a gate drive of low frequency asynchronous bursts at max duty cycle as the zero power comparator comes on and off depending only on the value of the error amp output.
    I don't think your issue is related to the hysteresis of the zero power comparator and I don't know why you see some MHz oscillation on the gate drive.
    Can you post the schematic and a scope waveform ?
    Regards
    John
  • 0 in reply to John Griffin1
    Prodigy 100 points

    Hi John,

    in the appached I show you what I find at pin 16. The first is one pulse in n load conditions, the second is the same but 500ns/div

    The schematics is the classic in literature. What do you suggest ?

    Thanks

    Best Regards

    gabriele

  • 0 in reply to Gabriele Montecchi
    Prodigy 100 points

    In the followings, the schematics of the circuits

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 34465 points

    Hi Gabriele,

    Can you please attache the schematic ?

    Click on  the  "Insert Code, Attach Files and more....." on the link below

    The locate the "insert file " and attach your schematic pdf.

    The high frequency pulses are not related to the no load condition but the low frequency burst is .

    What is your switching frequency ?

    If I can view your schematic it will help a lot

    Regards

    John

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 34465 points
    I don't see anything wrong in how the schematic is wired up but I do notice that the gate driver ic is a TC4428 dual driver that is used for some other signal that is not associated with the pfc. Can you disconnect this signal while you test the pfc stage ?
    I just wonder where the source of this high frequency is coming from since your switching frequency is set at 46kHz.
    Is the pcb layout good ?
    The voltage loop pole cap (C60) could be increased from 10nF to 150nF and I would try changing the zero capacitor (C57) from 470nF to 2.2uF.

    Also I would add a 100 Ohm resistor from the DRV pin (pin 16) to the TC4428 input in case there is some interaction between the high impedance input of the TC4428 and the low impedance output of the UCC3818.

    Regards
    John
  • 0 in reply to John Griffin1
    Prodigy 100 points

    Hi John,

    I tried 100ohm between pin 16 and the driver with no results. But according with slua309c application report, I added a 3.3Mohm between ref and pin 4 solving the problem.

    What do you think about?

    Regards

    gabriele

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 34465 points
    Hi Gabriele,
    I do not agree with the assumptions made in SLUA309C.
    If you wish to keep the pfc going out of burst mode then you need to keep the error amp output above 0.33V.
    You can do this by decreasing the current into IAC and I will explain to you how this works:
    The output of the multiplier sets the power signal from the pfc.
    At light load you can have the correct multiplier output by adjusting the two multiplier inputs : EA output and IAC input.
    So if you lower the IAC current signal then the closed loop nature of the pfc will cause the EA output to rise.
    If you lower IAC just enough then you can have EA out always above 0.33V over the full load range.
    This means that you will have very narrow pulses at light load instead of the burst type of switching.
    So you can experiment with increasing the resistors from IAC up to the high voltage input.
    The EA output needs to be within its linear range and it is a maximum at low line full power and a minimum (0.35v say) at min load high line)

    This will give you a fixed switching frequency but it does not explain the high frequency pulses on your scope plots.
    These may disappear and not be an issue if you adjust the IAC resistors but you may want to look at the suggestions I offered in my previous response.

    Regards
    John
  • 0 in reply to John Griffin1
    Prodigy 100 points

    Hi John,

    I appreciate a lot your suggestions and I tried to increase the resistors from IAC in order to lower the IAC current and to increase the EA output, but without any results. I can say that with another circuit and with another UCC3818, I found no problems without load because the voltage on pin 7 VAOUT is 0.8V. 

    Best Regards

    gabriele

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 34465 points
    Hi Gabriele,
    I understand. The only solution is to keep the EA output above the zero power threshold.
    You can only do this by decreasing the IAC input or by increasing the minimum load.
    May I ask what is different about this other circuit ? The one that has no issue.

    Regards
    John
  • 0 in reply to John Griffin1
    Prodigy 100 points

    Hi John,

    I agreed with you that the best solution is to keep the EA higher than power zero threshold. The second circuit I tested ( without any issues ), has the same components with theirs tolerances.

    Regards

    gabriele

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 34465 points
    Hi Gabriele,
    If the circuit is the same then increasing value of the IAC resistors should increase the EA output above the no load power threshold.
    I think you should pursue this approach until you can get EA output above 0.33V
  • 0 in reply to John Griffin1
    Prodigy 100 points

    OK John,

    I try to do what you suggest in the previous one and in other circuits. Another test I want to carry on if the replacement of the UCC3818 only.

    Regards.

    gabriele

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 31580 points
    Hi Gabriele,

    Do you have an update ?

    Thanks

    Peter
  • 0 in reply to Peter Meaney
    Prodigy 100 points

    Hi Peter,

    regarding the issue at light load in UCC3818, I found n°3 boost circuits working correctly and only one with the problem of high frequency at the output. At this point I suppose something wrong in the driver TC4428 tied at the output and I'll try to investigate in this direction.

    Thank

    gabriele

  • 0 in reply to Gabriele Montecchi
    TI__Mastermind 31580 points
    Hi Gabriele,

    Can we close this post and open a new one if you have more questions?

    Regards

    Peter
  • 0 in reply to Peter Meaney
    Prodigy 100 points
    Hi Peter,
    we can close this post and I'll keep this in mind for future questions.
    Thanks a lot.
    gabriele